ABSTRACT
Regionalized variable theory has been much used recently for making quantitative maps of single or multiple soil properties. The theory provides a convenient means of summarizing soil spatial variability in the form of an auto (or single property) variogram that can be used to estimate weights for interpolating the value of a given soil property at an unsampled location. Co-kriging is the logical extension of kriging to situations where two or more variables are spatially interdependent and the one of immediate interest is undersampled. Co-kriging could be useful for mapping a property that is expensive to measure by making use of a large spatial correlation between the property of interest and some cheaper-to-measure attribute. Elevation data can be used to advantage in a co-kriging procedure for mapping alluvial topsoils polluted with zinc in the floodplain of the Geul. Because elevation data are readily available on the altitude map of the Netherlands Topographic Survey the improvements cost very little additional expense.
